NPCL: Neural Processes for Uncertainty-Aware Continual Learning
Date
2023
Authors
Jha, S.
Zhao, H.
Gong, D.
Yao, L.
Editors
Advisors
Journal Title
Journal ISSN
Volume Title
Type:
Conference paper
Citation
Proceedings of the 37th Annual Conference on Neural Information Processing Systems (NeurIPS 2023), as published in Advances in Neural Information Processing Systems, 2023, vol.36, pp.34329-34353
Statement of Responsibility
Saurav Jha, Dong Gong, He Zhao, Lina Yao
Conference Name
37th Annual Conference on Neural Information Processing Systems (NeurIPS) (10 Dec 2023 - 16 Dec 2023 : New Orleans, Louisiana, USA)
Abstract
Continual learning (CL) aims to train deep neural networks efficiently on streaming data while limiting the forgetting caused by new tasks. However, learning transferable knowledge with less interference between tasks is difficult, and real-world deployment of CL models is limited by their inability to measure predictive uncertainties. To address these issues, we propose handling CL tasks with neural processes (NPs), a class of meta-learners that encode different tasks into probabilistic distributions over functions all while providing reliable uncertainty estimates. Specifically, we propose an NP-based CL approach (NPCL) with task-specific modules arranged in a hierarchical latent variable model. We tailor regularizers on the learned latent distributions to alleviate forgetting. The uncertainty estimation capabilities of the NPCL can also be used to handle the task head/module inference challenge in CL. Our experiments show that the NPCL outperforms previous CL approaches. We validate the effectiveness of uncertainty estimation in the NPCL for identifying novel data and evaluating instance-level model confidence. Code is available at https://github.com/srvCodes/NPCL.
School/Discipline
Dissertation Note
Provenance
Description
Access Status
Rights
© 2023 Neural Information Processing Systems Foundation, Inc.